1. Field of Invention
The invention relates to smoke detection, and more specifically to a method and apparatus for calibrating a smoke detector prior to installation and for testing the operation of the detector after installation and use.
2. Description of the Prior Art
Prior art smoke detectors typically include a dark chamber through which airborne particles of smoke are free to circulate. A source of light, such as an infrared emitter, directs illumination along a defined path extending into the chamber. A photoelectric sensor is positioned out of the path of direct illumination, but is aimed to view the chamber and illumination scattered or reflected from the path by circulating particles including smoke. When the sensor detects a level of illumination above a predetermined threshold, it issues an alarm signal.
Smoke detectors may be calibrated prior to installation and monitored for proper performance throughout their useful life. During calibration, an atmosphere representing a predetermined level of obscuration, such as three percent per foot, may be injected into the chamber and the smoke detector adjusted to alarm at the resulting signal level. The calibration level is chosen to represent the conditions that would exist when a fire is in its early stages of development.
Monitoring the detector after installation is somewhat more difficult. Frequently the detector must be removed from its installed location so it can be tested in a manor similar to that used prior to installation. Still, a satisfactory solution is not so simple. Detectors accumulate dust and other reflecting material in their chambers over time. The dust reduces the amount of obscuration required to activate an alarm, increasing the sensitivity of the detector and its tendency toward false alarms. Other problems may occur with opposite effects. A bug or foreign matter may partially block the source of illumination, decreasing the sensitivity of the detector and its ability for early warning. In either case the detector may still have an extended period of useful life, but its sensitivity and remaining life are difficult to determine with calibration samples.
Statistical sampling has been employed to estimate changes in detector performance. Many variables are involved, however, because the characteristics of the individual detector are seldom retained after installation. Each detector is different from other detectors in the same family, and, of course, the conditions of installation vary greatly. As noted above, some effects tend to increase sensitivity while other effects may reduce sensitivity, and, although not entirely random, historical changes are very difficult to predict.